I just want to add few things (if they matter): -I am running both servers on VMWare ESX -nothing else is running on servers (even heartbeat is down) -network connectivity works fine, I can connect to ssh of problematic server
On Fri, Feb 6, 2009 at 12:16 PM, Dusan Djordjevic <[email protected]> wrote: > Hi all, > > I am trying to install two LVS servers connected with Heartbeat. As > first step i installed just two servers and tested if pure LVS is > working. Strange, with identical setup, it works on one server, not on > another. > > Here is setup: > > both nodes are ubuntu 8.04 > ipvsadm is installed from standard ubuntu repo (apt-get install ipvsadm) > > I added following setup to /etc/rc.local > > echo 1 > /proc/sys/net/ipv4/ip_forward > echo 0 > /proc/sys/net/ipv4/conf/all/send_redirects > echo 0 > /proc/sys/net/ipv4/conf/eth0/send_redirects > > From both LVS, I can reach real servers. > > Here is LVS setup I am testing: > > ipvsadm -C > ipvsadm -A -t 192.168.21.100:80 -s rr -p 900 > ipvsadm -a -t 192.168.21.100:80 -r 192.168.20.1 -m -w 100 > ipvsadm -a -t 192.168.21.100:80 -r 192.168.20.2 -m -w 100 > ipvsadm -a -t 192.168.21.100:80 -r 192.168.20.3 -m -w 100 > > On one of nodes, it works withou any problem. On other one, it is not > working. Here is output from tcpdump (I am trying to connect to lvs on > address 192.168.21.100 from server 192.168.21.10). > > r...@ucplb-second-node:~# tcpdump -ln host 192.168.21.10 > tcpdump: verbose output suppressed, use -v or -vv for full protocol decode > listening on eth0, link-type EN10MB (Ethernet), capture size 96 bytes > 06:13:54.197763 arp who-has 192.168.21.100 tell 192.168.21.10 > 06:13:54.198226 arp reply 192.168.21.100 is-at 00:50:56:92:4d:a2 > 06:13:54.197809 IP 192.168.21.10.46204 > 192.168.21.100.80: S > 3085705580:3085705580(0) win 5840 <mss 1460,sackOK,timestamp 1070655 > 0,nop,wscale 5> > 06:13:54.197867 IP 192.168.21.10.46204 > 192.168.20.3.80: S > 3085705580:3085705580(0) win 5840 <mss 1460,sackOK,timestamp 1070655 > 0,nop,wscale 5> > 06:13:57.186597 IP 192.168.21.10.46204 > 192.168.21.100.80: S > 3085705580:3085705580(0) win 5840 <mss 1460,sackOK,timestamp 1070955 > 0,nop,wscale 5> > 06:13:57.186645 IP 192.168.21.10.46204 > 192.168.20.3.80: S > 3085705580:3085705580(0) win 5840 <mss 1460,sackOK,timestamp 1070955 > 0,nop,wscale 5> > 06:14:03.186451 IP 192.168.21.10.46204 > 192.168.21.100.80: S > 3085705580:3085705580(0) win 5840 <mss 1460,sackOK,timestamp 1071555 > 0,nop,wscale 5> > 06:14:03.186489 IP 192.168.21.10.46204 > 192.168.20.3.80: S > 3085705580:3085705580(0) win 5840 <mss 1460,sackOK,timestamp 1071555 > 0,nop,wscale 5> > > Any idea what might cause the problem. I am very confused since I > literally copied configuration from another node where it works. Any > idea how to troubleshoot this ? > > Thank you in advance. > > Dusan > -- Dusan Djordjevic, CMO | Troxo doo | http://www.troxo.com tel +381 18 29 29 40 | fax +381 18 526 917 | mob +381 69 22 44 617 _______________________________________________ Please read the documentation before posting - it's available at: http://www.linuxvirtualserver.org/ LinuxVirtualServer.org mailing list - [email protected] Send requests to [email protected] or go to http://lists.graemef.net/mailman/listinfo/lvs-users
